Fluctuations in benthic foraminiferal faunas over the last 130,000 yr in four piston cores from the Norwegian Sea are correlated with the standard worldwide oxygen-isotope stratigraphy. One species, Cibicides wuellerstorfi, dominates in the Holocene section of each core, but alternates downcore with Oridorsalis tener, a species dominant today only in the deepest part of the basin. O. tener is the most abundant species throughout the entire basin during periods of particularly cold climate when the Norwegian Sea presumably was ice covered year round and surface productivity lowered. Portions of isotope Stages 6, 3, and 2 are barren of benthic foraminifera; this is probably due to lowered benthic productivity, perhaps combined with dilution by ice-rafted sediment; there is no evidence that the Norwegian Sea became azoic. The Holocene and Substage 5e (the last interglacial) are similar faunally. This similarity, combined with other evidence, supports the presumption that the Norwegian Sea was a source of dense overflows into the North Atlantic during Substage 5e as it is today. Oxygen-isotope analyses of benthic foraminifera indicate that Norwegian Sea bottom waters warmer than they are today from Substage 5d to Stage 2, with the possible exception of Substage 5a. These data show that the glacial Norwegian Sea was not a sink for dense surface water, as it is now, and thus it was not a source of deep-water overflows. The benthic foraminiferal populations of the deep Norwegian Sea seem at least as responsive to near-surface conditions, such as sea-ice cover, as they are to fluctuations in the hydrography of the deep water. Benthic foraminiferal evidence from the Norwegian Sea is insufficient in itself to establish whether or not the basin was a source of overflows into the North Atlantic at any time between the Substage 5e/5d boundary at 115,000 yr B.P. and the Holocene.